Happy Labor Day! Regularly Scheduled Articles Will Resume on Tuesday, September 2.
March 18, 2010
Brendan Ryan Afield
I'm fond of saying that fielding metrics are probably best used in a wisdom-of-crowds kind of way, where you should look at all or as many of them as possible for any one player, and take the various and sometimes differing interpretations of his performance as a collective broad hint as far as a player's ability. Taking any one of them as gospel is a matter of choice, of course, but since all of the systems have their own quirks, virtues, and weaknesses, I favor casting a wide net.
The reason I bring this up is the interesting case of Brendan Ryan's fielding prowess. As I wrote a bit in this year's Cardinals essay in the Baseball Prospectus 2010, Skip Schumaker's conversion to second base wasn't exactly pretty; it worked, but it came with a cost in the field. But across the bag, Schumaker (and Cardinals pitchers) seemed to get the benefit of a huge defensive contribution from a new regular shortstop--not Khalil Greene, as they might have anticipated after trading for him, but organizational soldier Brendan Ryan. Using just FRAA, I couldn't help but be impressed by what Clay's Fielding Runs had to say in its evaluation of Ryan, crediting him with 31 runs of value. That's an enormous number, so with Eric Seidman's help, I took it up a notch, using FRAA2, the flavor of Fielding Runs that adjusts for league difficulty. Using Clay's metrics gives us the advantage of seeing everybody ever, putting Ryan's season into historical context by relying on the records at our disposal:
I stopped with Vizquel for a top 10, but in point of fact he's tied with Frank Crosetti's 1939 season with the Yankees and Art Fletcher's 1917 season with the Giants with a FRAA2 of 32. OK, that's pretty impressive as a group. It's not a rate metric, of course, but it does give us a broad-strokes comment on some great-fielding shortstops playing regularly. A couple of regulars from pennant winners, but a few from losing ballclubs as well. A few names we recognize and are usually associated with defensive excellence--Everett, Vizquel, Tulo, but also Peckinpaugh--but already, at this early juncture, there a few surprises. Tommy Thevenow? Wow, a positive thing to say about the man perhaps best known for having to play short for the 1930 Phillies, a team infamous for its ability to allow runs, not prevent them.
To see what other tools in the box have to say about Ryan, for a similar metric that, like Clay's metrics, relies on play-by-play data, I referred to Sean Smith's Total Zone, which you'll find for individual players on the incomparable Baseball-Reference (for more info, you can surf here). What does Total Zone's Runs Above Average say about the single-season all-time greats? Well, it's not quite so eager to propel Ryan to such heights, producing this top 10:
That gives a few of the same names in terms of further back in history (Thevenow, Fletcher, and Turner), gives Mark Belanger and Ozzie Smith some props that certainly feel good on a certain visceral level for those of us willing to admit we're fortysomethings, and provides a number that provides some support to the Mets' decision to play Rey Ordonez regularly. It also ends up agreeing that Adam Everett's 2006 season deserves tremendous regard. And Ryan? He's tied for 122nd all-time with 16 runs above average, tied with Tulo's 2009 campaign, Royce Clayton's 1995, or Bucky Dent's 1978, among others. That's good, handy, lovely... but also not quite epic.
Turn to other metrics, and Ryan fares well. Via John Dewan's Plus/Minus, it looks like Ryan's 2009 was the second-best shortstop season in MLB in terms of plays (+25) and runs (+19), rating behind the slick-fielding Jack Wilson in both regards. Obviously, that's not historic, and Plus/Minus doesn't cover the full spread of history. And UZR? It ranks him fifth in raw valuation (+10.6), again, saying he's an extremely valuable asset in the field.
Intrigued by the results, I couldn't help but wonder about a few of the factors in play: Joel Pineiro's even more pronounced grounderifficness? Some form of compensation for the recently converted Schumaker? The benefit of playing behind two of the game's best pitchers, Chris Carpenter and Adam Wainwright, since their form of dominance might lead to fewer well-hit balls in play? Musing about this stuff on BP's internal distribution list, I ended up pestering Russell Carleton about his thoughts on Ryan, especially since Russell's got his own fielding tool, OPA! Russell spun his system off of what Dan Fox had started at BP with Simple Fielding Runs as well as Sean Smith's work. But Russell brought into play his own insights into fielding interpretation. As he explains it:
My hook was that I broke things down into component parts. For example, on a ground ball, I thought it was foolish only to count completed plays. For example, a grounder to short, Theriot gets to it, fields it, and throws it to first, but Derek Lee drops the ball. Theriot did everything right... he should get credit, and Lee get the blame. So, I broke grounders down to range, hands, arm, and catch (over at first, mostly). The idea is that if a shortstop gets to the ball, then 70-something percent of the time (probably more, I'm just making numbers up), there will be an out recorded. If he fields it cleanly, that jumps to 90 percent, if he makes the throw, it's 99.5 percent...
All of which is pretty cool to have dug up, and I'm grateful to fellow statheads beyond BP, as well as Eric, Clay, Russell, and Jay Jaffe for their input within our own ranks. As an exercise in casting a broad net to learn something general about a single player's value in the field, there was certainly broad agreement: Ryan's a valuable defender. I know that's perhaps obvious, but basically I wanted to see what all of the systems said about a player which one of them singled out and described as remarkably good. In the broad strokes, doing so settles for me, the liberal arts type, the proposition of roughly where we should place Ryan's greatness in the field: among the ranks of the good to really good, given that the broad consensus seems to favor a more conservative interpretation of his value.